안녕하세요, 경인교육대학교 영어교육과 교수 성민창입니다. 저의 연구 관심 분야는 영문법교육 및 습득, 인지언어학, 코퍼스분석, 인공지능 챗봇입니다. 연락처는 mcsung(at)ginue.ac.kr 또는 minchang.sung(at)gmail.com 입니다.
Hello, this is Min-Chang Sung, Associate Professor at Department of English Education, Gyeongin National University of Education, S. Korea. My research interests include English grammar education and acquisition, cognitive linguistics, corpus analysis, and AI chatbot. You can contact me via mcsung(at)ginue.ac.kr or minchang.sung(at)gmail.com.
English textbooks for Korean elementary school students
- I belong to a team of professors, teachers, publishers, and IT professionals who develop English textbooks for Korean Elementary School Students.
- We have already developed textbooks for Grade 3 and 4 and passed the MOE's Textbook authorization and review. Now we are working on textbooks for Grade 5 and 6. 🤞
Development of an automatized speaking assessment system for Korean learners of English
- English speaking is something every Korean learner is interested in but struggles with. A research team of four wonderful professors and staff is addressing this issue by developing a new speaking assessment system. A beta version is available here.
Effects of visual cues on L2 English sentence processing
- Pictures are helpful in L2 reading, but we know little about their specific effects on L2 sentence processing. In this project, a team of three researchers has found that pictures accelerate L2 sentence processing. This effect is being analyzed with different types of sentences.
AI writing tutor chatbots for L2 writers
- This project addresses the following questions:
a. How can we develop AI writing tutor chatbots to promote L2 process writing and prevent side effects such as over-reliance?
b. What affordances do the tutor chatbots provide for L2 writers?
c. What strategies do L2 writer use when interacting with the tutor chatbots?
A GPT-API-based feedback system for pre-service teachers' lesson planning
- In my courses, I spend a LOT of time giving feedback on pre-service teachers' lesson plans and, of course, often fail to provide immediate feedback mostly due to time constraints. Other professors struggle with similar challenges.
- So, Dr. Jo and I developed an automatic feedback system using GPT-API. This system provides "literally" immediate feedback to pre-service teachers who write their lesson plans in its template.
- We have already published a paper focusing on the development process, and more will come, hopefully, to discuss its feedback quality, its effects on lesson plan revision, and user responses.
Comparing human proofreading vs. LLM proofreading
- In the past, L2 English writers mostly relied on professional human proofreaders to improve their writing, but now, many L2 writers are using LLMs such as ChatGPT as a major source of proofreading.
- Noting this shift, we decided to compare humans vs. LLMs in their proofreading performance, focusing on vocabulary and grammar improvements.
AI technology trends and their implications for Korea’s national English curriculum
- All the faculty members in my department are collaborating to publish a book that suggests directions for future revision of Korea's national English curriculum. As part of this project, I am writing a chapter that reviews AI technology trends and explores their Implications for curriculum revision.